Symptoms - scheduler experiment

This document analyses the self-developed scheduler component in Android-based mobile devices as part of the SyMptOMS project.

Reproduce paper

To create the PDF of the computational notebook you can run the following commands in a new R session. If you have problems rendering the PDF you can execute each chunk independently in RStudio.

This document does not install the required R packages by default. You can run the script install.R to install all required dependencies on a new R installation, or use install.packages(..) to install missing R packages.

The plots and tables use the packages ggplot2, knitr::kable() and kableExtra.

Required libraries and runtime environment description are as follows.

library(tidyverse)
library(kableExtra)
library(here)
library(googledrive)
library(lubridate)
library(stringr)
library(scales)
library(plotly)
library(grid)
library(gridBase)
library(gridExtra)

Data download

Data preparation

Key Variables:

• quantitative (interval)

  • plan_date (datetime): task planning time.
  • exec_date (datetime): task execution time.
    

• quantitative (ratio)

  • delay (numeric): execution delay in seconds ((exec_date - plan_date) - 60).

• categorital (ordinal)

  • battery(numeric): battery level

• categorical (nominal)

  • device_id/device_name: device identifier/name
  • exp_id: experiment identifier.
  • scheduler: scheduler type.

Scheduler assessment

Observations

• 129.922 observations registered in both experiments. Table above summarises ranges and total numbers per device.
• Nvidia Shield Tablet has stopped triggering alarms on 2020-04-17 22:27:06 when running the basic scheduler. Since there’s no watchdog in the basic version, this device is not going to trigger alarms anymore. Died!

Table. Descriptive data (total, ranges) per device.

exp_id	scheduler	device_name	# observations	start datetime	end_datetime	mean delay	sd	mean - 2sd	mean + 2sd
#1	Advanced	BQ Aquaris V	21391	2020-03-30 12:15:03	2020-04-14 13:07:57	0.703	2.431	-4.159	5.565
#1	Advanced	Nvidia Shield Tablet	21453	2020-03-30 12:15:03	2020-04-14 13:08:14	0.638	2.169	-3.700	4.976
#1	Advanced	Xiaomi Mi A1	21465	2020-03-30 12:23:27	2020-04-14 13:07:50	0.476	3.513	-6.550	7.502
#1	Basic	Honor 9	21574	2020-03-30 12:00:03	2020-04-14 13:01:35	0.351	2.456	-4.561	5.263
#1	Basic	Motorola Moto G	21619	2020-03-30 12:00:07	2020-04-14 13:01:16	1.014	2.755	-4.496	6.524
#2	Advanced	Honor 9	5783	2020-04-15 11:00:02	2020-04-19 11:59:14	0.351	2.456	-4.561	5.263
#2	Advanced	Motorola Moto G	4249	2020-04-16 12:09:53	2020-04-19 12:00:11	1.014	2.755	-4.496	6.524
#2	Basic	BQ Aquaris V	4429	2020-04-15 11:16:26	2020-04-18 13:30:10	0.703	2.431	-4.159	5.565
#2	Basic	Nvidia Shield Tablet	3528	2020-04-15 11:16:18	2020-04-17 22:27:06	0.638	2.169	-3.700	4.976
#2	Basic	Xiaomi Mi A1	4431	2020-04-15 11:16:23	2020-04-18 13:32:22	0.476	3.513	-6.550	7.502

Outliers

Observations:

• 1.6% (2.108/129.922) of outliers. Outlier breakdown below.

• Basic - Nvidia Shield Tablet does not show outliers!

• uneven distrintio of outliers per device!

Table. Outliers total per sheduler/device.

scheduler	device_name	# outliers
Advanced	BQ Aquaris V	65
Advanced	Honor 9	54
Basic	Honor 9	6
Advanced	Motorola Moto G	461
Basic	Motorola Moto G	1079
Advanced	Nvidia Shield Tablet	231
Advanced	Xiaomi Mi A1	210
Basic	Xiaomi Mi A1	2

Distribution of delay over plan_date

See the shiny app!

Distribution of delay over battery (without outliers)

Boxplots

Basic statistics applied to delay and device_name. Given a device, how do the box plots change on each scheduler?

## Warning in RColorBrewer::brewer.pal(N, "Set2"): minimal value for n is 3, returning requested palette with 3 different levels

## Warning in RColorBrewer::brewer.pal(N, "Set2"): minimal value for n is 3, returning requested palette with 3 different levels